1 paper
Qinshi Zhang, Weipeng Deng, Zhihan Jiang +4
In model-based learning, the agent learns behaviors by simulating trajectories based on world model predictions. Standard world models typically learn a stationary transition funct…